Agmatine, a product of arginine decarboxylation in mammalian cells, is believed to govern cell polyamines by inducing antizyme, which in turn suppresses ornithine decarboxylase (ODC) activity and polyamine uptake. However, since agmatine is structurally similar to the polyamines, it is possible that it exerts antizyme-independent actions on polyamine regulatory pathways. The present study determined whether agmatine inhibited ODC activity and polyamine transport in rat pulmonary artery endothelial cells (PAECs) by an antizyme-dependent mechanism. Agmatine caused time-dependent reductions in ODC activity, which occurred before increases in antizyme. Interventions that suppressed proteosome function caused large increases in ODC activity but failed to attenuate inhibitory effects of agmatine. When agmatine was present in the culture medium, 14C-polyamine uptake was competitively inhibited as evidenced by substantial elevations inK m values. If PAECs were incubated with agmatine for periods sufficient to increase antizyme, there were modest decreases in V max for putrescine and spermidine but not for spermine. These effects of agmatine on polyamine transport were insensitive to protein synthesis inhibition. Collectively, our findings show that agmatine decreases ODC activity and polyamine transport in PAECs, but a causal role for antizyme in these actions of agmatine is difficult to establish. Nevertheless, these observations are consistent with a model in which PAECs express both antizyme-1 and -2, but only the latter contributes to agmatine-mediated suppression of ODC activity.
Agmatine, formed by the decarboxylation of arginine via the enzyme arginine decarboxylase, is widely and unevenly distributed in mammalian tissues and is present in detectable amounts in plasma (Raasch et al., 1995;Feng et al., 1997). Once believed to be solely an intermediate in bacterial polyamine metabolism, experiments in mammals have indicated that agmatine serves as a neurotransmitter by acting as an agonist at imidazoline and α2-adrenergic receptors (Li et al., 1994; Reis and Regunathan, 1998, 1999) and by modulatingN-methyl-d-aspartate channel activity (Yang and Reis, 1999). Agmatine also evokes a variety of other pharmacological effects, including stimulation or inhibition of nitric-oxide synthase activity (Auguet et al., 1995; Galea et al., 1996; Morrissey and Klahr, 1997; Schwartz et al., 1997), modulation of insulin release from pancreatic β cells (Shepherd et al., 1996; Chan, 1998), and inhibition of smooth muscle proliferation (Regunathan et al., 1996; Regunathan and Reis, 1997; Molderings and Gothert, 1999).
Along with the above-mentioned actions, agmatine has recently been shown to interact with pathways governing cell polyamines. The polyamines putrescine, spermidine, and spermine are low-molecular-weight organic cations required for cell growth and differentiation. Agmatine is believed to deplete cell polyamine contents by inducing antizyme (Satriano et al., 1998), which then inactivates the initial enzyme in polyamine synthesis, ornithine decarboxylase (ODC), and promotes its 26S proteosome-dependent degradation (Murakami et al., 2000). The polyamine transport pathway also is inhibited by antizyme as evidenced by substantial reductions inV max (Mitchell et al., 1994). Nevertheless, antizyme-mediated actions of agmatine have not been observed in all cell types. In hepatocytes, agmatine induces spermidine/spermine acetyltransferase and inhibits ODC activity and polyamine import without involvement of antizyme (Vargiu et al., 1999).
To explore the possibility that some of agmatine's actions are independent of antizyme, we examined its effects on cultured pulmonary artery endothelial cells (PAECs). We focused our studies on endothelial cells because polyamine regulatory pathways have been reasonably well characterized in this cell type (Morgan, 1992; Sokol et al., 1993;Morrison and Seidel, 1995). Endothelial cells also are known to express arginine decarboxylase (Regunathan et al., 1996), thus suggesting that locally formed agmatine could modulate polyamine-dependent cell functions.
Materials and Methods
Rat Main Pulmonary Artery Endothelial Cell Cultures.
Rat main pulmonary artery endothelial cells were isolated and cultured as described previously (Babal et al., 2000). In brief, main pulmonary arteries were isolated from 250 to 300 g Sprague-Dawley rats killed with an overdose of pentobarbital sodium [Nembutal; Bergen-Brunswig (Drug Division), Mobile, AL]. Isolated arteries were opened and the intimal lining was carefully scraped with a scalpel. The harvested cells were then placed into flasks (Corning, Corning, NY) containing F-12 nutrient mixture and Dulbecco's modified Eagle's medium (DMEM) mixture (1:1) supplemented with 10% fetal bovine serum, 100 U/ml penicillin, and 0.1 mg/ml streptomycin (Life Technologies, Grand Island, NY). Culture medium was changed once per week and, after reaching confluence, the cells were harvested using a 0.05% solution of trypsin (Life Technologies). The endothelial cell phenotype, confirmed by acetylated low-density lipoprotein uptake, Factor VIII-RAg immunostaining, and the lack of immunostaining with smooth muscle cell α-actin antibodies (Sigma, St. Louis, MO), persisted for at least 15 passages.
Western Analysis of ODC-Antizyme.
The cells were seeded in six-well plastic plates at 5 × 105cells/well and, when confluence was attained, were treated with agmatine (Sigma) for 0.5, 2, and 24 h. Cells were then washed twice with phosphate-buffered saline (PBS), lysed in 3% SDS electrophoresis buffer, and 120 μg of protein was applied to an SDS/10% polyacrylamide gel. After separation, samples were transferred to nitrocellulose filters (Bio-Rad, Hercules, CA). Membranes were blocked in 5% nonfat dried milk in PBS with 0.05% Tween 20 and were incubated overnight at 4°C with primary, polyclonal antibody recognizing both antizyme-1 and antizyme-2, diluted 1:1000 in PBS-Tween 20 with 0.5% bovine serum albumin. Before applying to the membrane, the diluted antibody was filtered through a 0.45-μm nylon filter (Corning). After washing, the membranes were incubated with 1:20,000 diluted horseradish peroxidase-conjugated goat anti-mouse IgG (Bio-Rad) for 1 h at room temperature and then revealed by chemiluminescence with the SuperSignal West Pico kit (Pierce, Rockford, IL).
Ornithine decarboxylase activity in PAECs was evaluated as previously described (Harrod et al., 1996). ODC activity, quantified as the amount of14CO2 released from 0.5 μCi of l-[1-14C]ornithine (New England Nuclear, Boston, MA) during a 60-min incubation at 37°C, was normalized to cellular protein content as determined by the Bradford assay.
Cells were seeded in 12-well plastic plates (Corning) at 3 × 105cells/well and grown to confluence. Polyamine uptake was determined as described previously (Aziz et al., 1995; Babal et al., 2000). After incubation under the indicated conditions, cells were rinsed with serum-free DMEM after which 1 ml of DMEM was added to each well and the cells were allowed to acclimate for 30 min. Subsequently, [14C]putrescine, -spermidine, or -spermine (New England Nuclear) plus an appropriate amount of unlabeled polyamine was added to each well. Cells were then incubated for 30 min, a duration that preliminary experiments indicated was in the linear range of uptake. Nonspecific polyamine import was determined as the uptake occurring in cells incubated at 4°C. At the appropriate time, media containing residual 14C-polyamines was aspirated, cells were rinsed twice with cold PBS, and overlayed with 1.5 ml of PBS containing 0.5% SDS for 30 min. The cell lysates were then transferred to scintillation vials, mixed with 4 ml of scintillation cocktail (Beckman Instruments, Fullerton, CA), and radioactivity determined using a Beckman LS 6500 liquid scintillation counter. Polyamine uptake rates were expressed in terms of uptake per 1 million cells per minute. GraphPad Prism software was used to fit polyamine concentration-uptake rate data to the Michaelis-Menten equation and to derive values ofK m andV max (Motulsky and Ransnas, 1987).
Overexpression of Mutant ODC in PAECs.
Using previously reported methods (Clifford et al., 1995), rat PAECs were stably infected with the retroviral vectors pLXSN and pLOSN. pLXSN is a Maloney murine leukemia virus-derived retroviral vector containing the neomycin resistance gene driven by an internal simian virus 40 promoter to provide a G418 selection marker. pLOSN is identical except it encodes a mouse ODC protein truncated at the carboxyl terminus and lacking the first PEST sequence required for antizyme-induced, proteosome-mediated degradation of the enzyme (Ghoda et al., 1989). All experiments performed on PAECs with overexpressed ODC were matched with pLXSN-transfected cells and with wild-type PAECs. No differences were detected between the wild-type andneor -transfected cells and results from these two control groups are pooled for presentation.
Data are expressed as the mean ± standard error of the mean. Differences in K mand V max values between control and agmatine-treated cells were determined according to the approach described by Motulsky and Ransnas (1987). In other experiments, one-way analyses of variance combined with Newman-Keuls tests were used to detect time- or concentration-related differences. Values ofp less than 0.05 were taken as evidence of statistical significance.
Initial studies sought to verify that agmatine elevates antizyme in PAECs. The Western analysis shown in Fig.1A demonstrates that although PAECs treated with 1 mM agmatine for 30 min failed to exhibit changes in antizyme abundance, exposure for 2 h was associated with a modest elevation that was more pronounced with 24 h of agmatine treatment. Results of three such Western analyses were semiquantitated by densitometry and pooled for display in Fig. 1B. Within 2 h of agmatine exposure, antizyme was elevated almost 2-fold, and by 24 h there was nearly a 5-fold increase.
Having confirmed that agmatine elevated antizyme abundance in PAECs, we next evaluated its actions on ODC activity and ODC protein abundance. As shown in Fig. 2, 1 mM agmatine was associated with substantial reductions in ODC activity within 30 min of its addition to the culture medium. Longer durations of exposure were accompanied by more pronounced ODC inhibition.
To explore the role of the proteosome in the inhibitory effects of agmatine, we used the well characterized proteosome inhibitor lactacystin. As shown in Fig. 3, a 24-h incubation with lactacystin was associated with increased ODC activity in wild-type PAECs. More interestingly, the proteosome inhibitor failed to prevent the approximate 95% reduction caused by 2 h of agmatine treatment, despite the fact that this duration of exposure is accompanied by marked elevation in PAEC antizyme content. In a companion study using an identical experimental design, the effect of agmatine and lactacystin on ODC activity was examined in PAECs stably transfected with a truncated ODC mutant resistant to antizyme-induced, proteosome-mediated degradation. Here, too, the rationale was that if the impact of agmatine on ODC activity was related to antizyme and the attendant proteosome-mediated degradation, then the inhibitory effects should be suppressed in the transfected PAECs despite the agmatine-induced increase in antizyme content. As expected, we found that baseline ODC activity was higher in the truncated ODC mutant (38 ± 3 nmol/mg/h) than in wild-type and vector-transfected PAECs (0.7 ± .18 nmol/mg/h). In addition, data presented in Fig. 3 show that lactacystin failed to elevate ODC activity in the ODC mutants as it did in wild-type cells; this outcome was anticipated since the truncated ODC in these cells is not degraded by a lactacystin-sensitive pathway. Finally, the inhibitory effect of agmatine on mutant ODC activity in the PAEC transfectants persisted to an extent similar to that observed in the wild-type and vector-transfected cells.
To determine whether agmatine inhibited polyamine uptake in PAECs, concentration-uptake rate curves were generated for the three14C-polyamines in the absence and immediately after addition of 1 mM agmatine to the culture medium. Companion studies determined polyamine uptake kinetics in agmatine-free medium after 2- or 24-h incubations with 1 mM agmatine. This latter protocol was used to examine agmatine's effects under conditions when antizyme was elevated but when potential competition between agmatine and polyamines for uptake pathways was minimized. The curves so generated, displayed in Fig. 4, and values forK m andV max presented in Table1, show that when agmatine was present in the culture medium, there were large increases inK m values for the uptake of all three polyamines with no changes in V max. In contrast, when cells were incubated with agmatine for 2 or 24 h after which 14C-polyamine uptake was determined in agmatine-free medium, the V maxvalues for putrescine and spermidine transport were reduced, whereas the K m value for putrescine was elevated. Protracted treatment with agmatine failed to alter [14C]spermine transport.
Although the acute effects of agmatine on polyamine uptake can most likely be ascribed to competition between agmatine and polyamines for the transporter (vide infra), the modest reduction inV max noted after 2 or 24 h of agmatine pretreatment was associated with elevated antizyme levels. To explore the possibility that transport inhibition under these conditions was antizyme-dependent, we determined whether the impact of agmatine persisted when antizyme was eliminated by protein synthesis inhibition (Mitchell et al., 1992). In preliminary studies, we examined the time-dependent effects of cycloheximide on baseline uptake of 0.3 μM of each 14C-polyamine. Peak increases occurred after 6 h of treatment followed by a gradual decline to values below baseline by 24 h (data not shown). This biphasic time course reflects the initial relief of antizyme-mediated repression of the transporter followed by a probable reduction in transporter synthesis or activity (Mitchell et al., 1992, 1994). Antizyme abundance, already at the lower limit of detection by Western analysis in control PAECs, was nondetectable after 6 h of cycloheximide (data not shown). In the experiments presented here, 1 mM agmatine was added to the culture medium after 6 h of treatment with cycloheximide and putrescine transport determined immediately thereafter. In a second protocol, agmatine was added for the last 2 h of cycloheximide treatment after which [14C]putrescine uptake was determined in agmatine-free culture medium. Only putrescine uptake was examined since the effect of 2 h of agmatine exposure was most prominent for this polyamine. As shown in Fig. 5, a 6-h incubation with cycloheximide caused substantial increases in PAEC putrescine uptake. However, the inhibitory effects of both the acute addition of agmatine to the culture medium or the more protracted treatment regimen to elevate antizyme abundance persisted in the presence of cycloheximide.
Understanding mechanisms by which agmatine governs polyamine regulatory pathways in ECs is important for several reasons. First, ECs express arginine decarboxylase (Regunathan et al., 1996) and may serve as a source of agmatine. Locally synthesized agmatine could thus act as an autocrine regulator of polyamine disposition in target ECs. Polyamines play an important role in signal transduction and may govern EC responses in vascular injury and repair (Olson et al., 1984;Atkinson et al., 1987). And, as recently pointed out by Satriano et al. (1999), agmatine may be useful for pharmacological manipulation of polyamine-dependent cellular events.
The initial suggestion that agmatine promoted antizyme-dependent inhibition of ODC activity was based on findings that agmatine induced time-dependent increases in antizyme abundance and that cytoplasmic extracts from agmatine-treated cells caused antizyme-mediated decreases in ODC activity in a cell-free bioassay system (Satriano et al., 1998). Since this initial report, it has become better appreciated that there is a family of mammalian ODC antizymes. Searches of genomic databases have revealed five nonallelic antizyme isoforms (Ivanov et al., 2000a); some information is available about three of its putative members. Antizyme-1 has been shown definitively to promote both ODC inactivation and degradation as well as inhibit polyamine uptake (Murakami et al., 2000). The initial characterization of antizyme-2 indicates that it shares antizyme-1's ability to bind and promoteinactivation of the enzyme as well as inhibit polyamine transport, but there is uncertainty regarding whether antizyme-2 promotes degradation of the ODC protein (Zhu et al., 1999). Expression of antizyme-3 seems to be confined to the testes where it regulates spermatogenesis (Ivanov et al., 2000b). Like the two members of the antizyme family noted above, antizyme-3 inhibits ODC activity. Little else is known of its biological properties.
Similar to its actions in other cell types, we found that agmatine increases the PAEC content of antizyme and inhibits both ODC activity and polyamine transport. Counter to our expectation, several aspects of the current data make it difficult to rigorously assign a causal role to any single antizyme in agmatine's effects on PAEC polyamine regulation. First, time courses for agmatine-induced effects on antizyme and inhibition of ODC were discordant; agmatine-mediated inhibition of ODC activity occurred well before the first detectable increase in antizyme. This could be explained by a lack of sensitivity of the Western analysis of antizyme, but it is noteworthy that although antizyme abundance in the baseline state was at the limits of detection, it was biologically significant as evidenced by the marked increases in ODC activity with lactacystin treatment and increases in polyamine transport caused by cycloheximide.
A second consideration obscuring the association between agmatine, antizyme, and ODC activity is that the inhibitory effects of agmatine were not suppressed when proteosome function was blocked by lactacystin or by use of a truncated, degradation-resistant ODC mutant. As mentioned previously, in vitro studies suggest that antizyme-2 may not promote ODC degradation in a cell-free system, although this has not been confirmed in mammalian cells (Zhu et al., 1999). Inasmuch as the antibody used in our experiments detects both antizyme-1 and -2, it is tempting to speculate that in PAECs, antizyme-2 could participate in the actions of agmatine on ODC. Another possibility is that in PAECs with suppressed proteosome function, the agmatine-mediated increase in antizyme results in formation of a stable ODC-antizyme complex, perhaps because the lack of degradation prevents release of antizyme from the complex. As a consequence, ODC remains inhibited by antizyme.
A role for antizyme in agmatine's effects on polyamine transport in PAECs is also difficult to establish. As noted above, antizyme-1 and -2 are known to repress polyamine transport; for antizyme-1, this inhibition is characterized by decreases inV max (Mitchell et al., 1994). Agmatine inhibited polyamine transport, but its actions were most pronounced when it was present in the culture medium and before detectable increases in antizyme. Under these conditions, our kinetic analysis showed that the prominent effects were onK m values for polyamine uptake, suggesting a competitive mechanism of action that differs from the reported antizyme-mediated effects onV max. When PAECs were exposed to agmatine for periods sufficient to elevate antizyme abundance and then transport was determined in the absence of agmatine in the medium, reductions in V max values for putrescine and spermidine uptake were modest and no effects were observed when spermine was the transporter substrate. Importantly, the agmatine-mediated suppression of putrescine uptake under these conditions was not blocked by protein synthesis inhibition, which has been shown previously to eliminate the contribution of antizyme (Mitchell et al., 1992, 1994).
A conservative interpretation of the present data is that interactions between agmatine, antizyme, and polyamine regulation are different in PAECs relative to the other cell types so far studied. The possibility cannot be excluded that agmatine-induced increases in antizyme are an epiphenomenon, and in this cell type antizyme plays no role in the effects of agmatine. However, one model that could explain our observations is that both antizyme-1 and -2 are operative in PAECs. The marked increases in ODC activity and polyamine transport caused by lactacystin and cycloheximide would be attributed to relief from the inhibitory effects of antizyme-1, whereas the inhibition of ODC activity by agmatine, which persists in the presence of lactacystin, is ascribed to antizyme-2. The apparent lack of antizyme involvement in agmatine's inhibition of polyamine transport, and perhaps ODC as well, could be linked to formation of a stable, nondegradable antizyme-ODC complex that obviates antizyme's inhibitory interactions with the polyamine regulatory pathways. As a consequence, the more prominent effects of agmatine on polyamine uptake are mediated by competition with polyamines at the transporter level. Studies on selected aspects of this model might prove useful for exploring the complex interactions between agmatine, antizyme, and polyamine regulatory pathways in PAECs.
In summary, results of the present studies show that agmatine profoundly inhibits ODC activity and polyamine uptake in PAECs. If antizyme(s) mediates these effects, its involvement is different from that in other cell types so far studied. Given the pathological role of polyamines in specific pulmonary disorders, notably monocrotaline- and hypoxia-induced pulmonary hypertension (Olson et al., 1984, 1986;Atkinson et al., 1987), the postulated importance of polyamine transport in long-term maintenance of vascular cell polyamine pools (Aziz et al., 1995), and the possibility that agmatine's effects on polyamine regulatory pathways could be exploited therapeutically, the mechanism of agmatine actions in this cell type warrants additional investigation.
- Received August 17, 2000.
- Accepted October 5, 2000.
Send reprint requests to: Mark N. Gillespie, Ph.D., Department of Pharmacology, College of Medicine, MSB 3130, University of South Alabama, Mobile, AL 36688. E-mail:
This investigation was supported in part by grants from the National Institutes of Health (HL36404 and HL38495).
- ornithine decarboxylase
- pulmonary artery endothelial cell
- Dulbecco's modified Eagle's medium
- phosphate-buffered saline
- endothelial cell
- The American Society for Pharmacology and Experimental Therapeutics